Image analysis – Image compression or coding – Interframe coding
Reexamination Certificate
1999-04-15
2003-07-15
Johns, Andrew W. (Department: 2621)
Image analysis
Image compression or coding
Interframe coding
C382S247000, C708S201000
Reexamination Certificate
active
06594396
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to adaptive difference computing elements and motion estimation apparatuses, and more particularly to an adaptive difference computing element and motion estimation apparatus dynamically adapting to input data with a reduced amount of circuit operation.
2. Description of the Background Art
A motion estimation apparatus is used in a moving picture compression system for MPEG (Moving Picture Experts Group) and performs a large number of computations. To date, various computation algorithms for the motion estimation apparatus have been proposed. For the moving picture compression system for MPEG, “VLSI Architectures for Video Compression-A Survey”, by P. Pirsch et al., Proc. IEEE Vol. 83, No. 2, pp. 220-246, 1995 and “ULSI Realization of MPEG
2
Realtime Video Encoder and Decoder-An Overview”, by M. Yoshimoto et al., IEICE Trans. Electron., Vol. E78-C, No. 12, pp. 1668-1681, 1995 are incorporated herein by reference.
In addition, for an LSI (Large Scale Integration) for motion estimation computation, “A Half-pel Precision MPEG
2
Motion-Estimation Processor with Concurrent Three-Vector Search”, by K. Ishihara et al., ISSCC Digest of Technical Papers, pp.1502-1509, 1995 and “A Motion Estimation Processor for MPEG
2
Video Real Time Encoding at Wide Search Range”, by A. Ohtani et al., Proc. IEEE Custom Integrated Circuits Conference, pp. 405-408, 1995 are incorporated herein by reference.
In the moving picture compression system for MPEG, difference must be calculated between pixel values (sample values) of a reference block and one of blocks to be searched corresponding to one of a large number of candidate vectors within a search range. Most of the computation performed by a whole system is the computation performed by the motion estimation apparatus. Therefore, it is critically important that difference calculation is performed by a circuit with power consumption which is as small as possible in order to achieve a moving picture compression system with reduced power consumption.
To cope with this subject, an encoding apparatus
220
has been proposed which is shown in FIG.
1
. Encoding apparatus
220
includes: a rounding circuit
201
rounding lower bits of a prescribed number of bits in accordance with an output from a quantization circuit
204
, which will later be described, for every sample of image data of a current frame which has been input from a video input; a frame memory
208
connected to an output of an addition circuit
212
, which will later be described, for storing image data; a rounding circuit
207
rounding lower bits of a prescribed number of bits in accordance with an output from a quantization circuit
204
which will later be described for every sample of image data which has been stored in frame memory
208
; a motion estimation computing element
202
receiving image data of the current frame input from the video input and that of a preceding frame stored in frame memory
208
respectively through rounding circuits
201
and
207
and calculating a sum of absolute difference between data elements of a reference block and those of one of blocks to be searched; a selector
211
connected to an output of motion estimation computing element
202
and the video input for selecting and outputting one of the above mentioned output or input; a discrete cosine translation circuit
203
connected to an output of selector
211
; a quantization circuit
204
connected to an output of discrete cosine translation circuit
203
for quantizing in accordance with a quantization control signal output from a output buffer portion
210
which will later be described; a variable length encoder
209
connected to outputs of quantization circuit
204
and motion estimation computing element
202
; an output buffer portion
210
connected to an output of variable length encoder
209
; an inverse quantization circuit
205
connected to an output of quantization circuit
204
; an inverse discrete cosine translation circuit
206
connected to an output of inverse quantization circuit
205
; and an addition circuit
212
connected to outputs of inverse discrete cosine translation circuit
206
and motion estimation computing element
202
for performing addition of data elements of two blocks and reconstructing the block.
In encoding apparatus
220
, an amount of data accumulated in output buffer portion
210
increases as the amount of encoded data increases. Thus, a quantization step value is increased to decrease the amount of encoded data. If the amount of encoded data decreases, conversely, it is controlled to increase. At the time, significant digit numbers of samples of the reference block and one of blocks to be searched applied to motion estimation computing element
202
are determined in accordance with an algorithm shown in FIG.
2
. NTB represents the number of non-trancated bits, and lower bits which are not included in the non-trancated upper bits are rounded in rounding circuits
201
and
207
. Quantization step values of the current and preceding frames are respectively represented by Qc and Qp. More specifically, if quantization step value Qc is equal to or smaller than quantization step value Qp and NTB is equal to or smaller than a possible maximum value 6, NTB is incremented by 1. Thus, the number of the lower bits to be rounded is decreased. Conversely, if quantization step value Qc is greater than quantization step value Qp and NTB is greater than 1, NTB is decremented by 1. Thus, the number of the lower bits to be rounded is increased. A number of significant bits of data computed by motion estimation computing element
202
can be decreased on the average in accordance with the algorithm. As described above, a method has been proposed which allows power consumption to be reduced by reducing calculation accuracy for adding absolute differences. Encoding apparatus
220
has been disclosed in “Reducing Hardware Complexity of Motion Estimation Algorithms Using Truncated Pixels”, IEEE ISCAS'97, 1997, by Zhongli He et al., which is herein incorporated by reference.
The above described encoding apparatus
220
suffers from a problem that the calculation accuracy of the motion estimation computing element is reduced as the lower bits are rounded.
In addition, if calculation is simply performed using data with all bits without rounding the lower bits, a signal change occurs many times which in turn increases power consumed by the circuit.
SUMMARY OF THE INVENTION
The present invention is made to solve the aforementioned problem. An object of the present invention is to provide an adaptive difference computing element and a motion estimation apparatus which consumes less power without any decrease in calculation accuracy.
Another object of the present invention is to provide an adaptive difference computing element and a motion estimation apparatus which can reduce the number of signal changes in a calculation circuit without any decrease in calculation accuracy.
Still another object of the present invention is to provide an adaptive difference computing element and a motion estimation apparatus capable of performing motion estimation without any decrease in calculation accuracy while substantially decreasing the number of bits of data to be calculated.
An additional object of the present invention is to provide an adaptive difference computing element and a motion estimation apparatus capable of performing motion estimation without rounding data with reduced power consumption.
Another additional object of the present invention is to provide an adaptive difference computing element and a motion estimation apparatus capable of performing motion estimation without any decrease in calculation accuracy by performing calculation for only a portion of data with reduced power consumption.
Another additional object of the present invention is to provide an adaptive difference computing element and a motion estimation apparatus capable of performing motion estimation without any decr
Moshnyaga Vasily
Yoshida Toyohiko
Dang Duy M.
Johns Andrew W.
LandOfFree
Adaptive difference computing element and motion estimation... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Adaptive difference computing element and motion estimation..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adaptive difference computing element and motion estimation... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3056187